Analysis: The "Lorenzo Jones" case emerges (SCOTUS Blog) [LWN.net]

Analysis: The "Lorenzo Jones" case emerges (SCOTUS Blog)

Posted Nov 9, 2009 23:35 UTC (Mon) by yokem_55 (subscriber, #10498)
Parent article: Analysis: The "Lorenzo Jones" case emerges (SCOTUS Blog)

From the oral argument transcript:

MR. STEWART: And all we've said is that it might be enough; that is, hard questions will arise down the road as to where do you draw the line, to what extent must the machine or the transformation be central --

CHIEF JUSTICE ROBERTS: So you think it's a hard question. If you develop a process that says look to the historical averages of oil consumption over a certain period and divide it by 2, that process would not be patentable. But if you say use a calculator, then it -- then it is?

I think we may have a friend in the Chief.

(Log in to post comments)

Analysis: The "Lorenzo Jones" case emerges (SCOTUS Blog)

Posted Nov 10, 2009 4:09 UTC (Tue) by gmaxwell (subscriber, #30048) [Link]

It's helpful to consider the discussion in the light of the USPTO guidelines: http://www.uspto.gov/patents/law/comments/2009-08-25_inte...

Though as an engineer I'm not sure how to apply the criteria given: If a patents claims regarding the programming of a general machine "is so abstract and sweeping that performing the process as claimed would cover substantially all practical applications of a judicial exception, such as a mathematical algorithm, the claim would not satisfy the test as the machine would not be sufficiently particular." and "A field-of-use limitation does not impose actual boundaries on the scope of the claimed invention."

Channelling the Knuth argument: All computer software is a mathematical algorithm, and if applying it to a particular field ("Fourier transforms, on a computer, FOR DETECTING PORN") isn't enough then I don't know what could be.

In the oral arguments both the attorneys and the justices seemed a little uninformed from a computer science perspective, for example they seemed to view executing a (possibly complex) series of operations on a calculator to be categorically different from programming a computer.

Often in law these kinds of issues are resolved by using special domain specific meanings of words (like 'computer' and 'mathematical algorithm') which an engineer wouldn't find especially agreeable, but the purpose of law isn't to be consistent with mathematics or engineering... if it wants to equate π=2.0 then it can. But in those cases it actually has to have its own somewhat consistent definition, and as far as I can tell it doesn't have one here.

In the oral arguments it sounded like the government conceded that they don't have a solution, but that they think it's reasonable to hammer one out over time by working on a case by case basis.

The resulting uncertainty can only be a boon to attorneys and a burden on everyone else.

Analysis: The "Lorenzo Jones" case emerges (SCOTUS Blog)

Posted Nov 11, 2009 0:34 UTC (Wed) by dmk (subscriber, #50141) [Link]

hm the patent defender argues, that the previous court took an arbitrary test and rejected the patent on that ground. and that that is not right. ? somehow a substantial argument is escaping me there.

the other party says that this isn't some big case where there shouldn't be any new ruling. it's just about confirming given practice and that the transform-or-machine test for ''processes'' is a valid test which doesnot preclude busniss patents, nor includes them and it is thus safe to use without creating havoc in industry-world.

i find it really weird. both sides say something, the court would love to finish off business patents, but no one wants to come up with a rule on how to exclude business-patents (aka software-patents?) without damaging the valid side of the patent system.

so to me they really seem to talk at cross purposes. but it looks as if that patent would be killed. but why it gets killed is the last open question.